TÉLUQ University

The effectiveness of Positive Behavioral Interventions and Supports (PBIS) in reducing major misbehavior has been demonstrated in many research studies. However, no research on the effects of PBIS on student behavior has been conducted in Quebec and other French-speaking regions. This study focuses on the results of PBIS implementation in Quebec schools in Canada. This article presents results following implementation in three secondary schools and one elementary school that recorded Office Discipline Referrals (ODRs) for the year preceding PBIS implementation and the following 3 years of PBIS implementation. An average annual rate of ODRs per student was calculated for each school and for each year. The results show a reduction in the average annual rate of ODRs per student in each of the 3 years following the implementation of PBIS. In total, there was a 78% reduction in the number of ODRs in Year 3 of PBIS across the four schools. The article also discusses the features of PBIS implementation in Quebec.

Under climate change, some forest ecosystems appear to be transitioning into net source of carbon dioxide (CO2), raising questions about the future role of soil respiration rate (Rs), which depends on hydroclimatic conditions. Conversely, well-drained forest soils could become more significant sinks of methane (CH4) under warming. The main objective of this study was to assess the effects of artificial soil warming on Rs and CH4 fluxes in a sugar maple forest at the northern limit of Quebec temperate deciduous forests in eastern Canada, and to evaluate the effect of species composition on soil response to warming. We measured Rs and CH4 fluxes during the snow-free period of 2021 and 2022 in 32 plots distributed across three forest types, half of which were artificially heated by approximately 2 °C with heating cables. Forest soils were a very consistent sink for CH4 and it did not respond to artificial soil warming nor was it sensitive to variations in soil moisture, ionic activity in soil solution and forest types. However, we observed an increase in Rs in response to warming in the heated plots, but only up to a threshold of about 15 °C, beyond which Rs started to slow down in respect to the control plots. We also observed a weakening of the exponential relationship between Rs and soil temperature beyond this threshold. This trend varied across the forest types, with hardwood-beech stands being more sensitive to warming than mixedwoods and other hardwoods. This greater response of hardwood-beech stands to warming resulted in a more significant downshift of Rs, starting from a colder temperature threshold, around 10–12 °C. This study highlights a potential plateauing of Rs despite rising soil temperature, at least in eastern Canada’s temperate deciduous forest, but this trend could vary from one forest type to another.

Teachers worldwide are expected to adapt to increasingly complex demands. Meanwhile, there is a shortage of qualified teachers in the profession. In this context, our paper explores the role of work intensification (WI) as a predictor of teacher turnover intention, an important antecedent that has never been explored amongst school teachers. The role of work–life conflict (WLC) is also considered, given the salience of this issue according to teacher unions. We distributed an online questionnaire to teachers from various sectors (preschool, primary, secondary, adult training, professional training and special education) through union listings and got 405 valid responses. We ran statistical analyses using PROCESS Macro v.4.2 for SPSS, and our results indicate a direct, significant and positive relationship between WI and intention to leave ( p ≥ 0.001; R ² = 0.179). Moreover, we found that WLC interacts with WI in its impact on intention to leave ( p ≥ 0.001; R ² = 0.191). Theoretical contributions are made using the job demands‐resources and conservation of resources theories, and practical implications for government and school leaders are discussed.

Political leaders are often regarded as the most qualified individuals to address modern societal challenges, owing to the knowledge they acquire through their experience in dealing with complex issues, governance and management, and working towards making impactful decisions. To understand the influence of prior knowledge on decision-making, we conducted a comparative analysis of complex decision-making performance in a politically themed computer-simulated microworld involving incumbent elected officials and a general population sample, each contrasted with a random-response baseline produced with randomly generated decisions. Participants were tasked to govern a country for re-election while maintaining financial stability. The pattern of results suggests that decision-making faces a ‘wall of complexity’ whether one is an elected official or a citizen. Although elected officials generally reported having greater political knowledge, their performance was still relatively poor. The elected officials and general population subgroups performed at the same level and only slightly better than chance. Addressing the societal challenges of our time requires elected officials to possess more than domain-specific prior knowledge.

Background Prominent theories of reading make the prediction that individual differences in children's word learning capacity determine the pace of their acquisition of reading skill. Despite the developmental nature of some of these theories, most empirical research to date has explored the relation between word learning capacity and reading at a single time point. The present study extends this research base by investigating whether earlier learning of the spelling and meaning of words is associated with later core aspects of reading: orthographic representations, word reading and reading comprehension. Methods Participants were 120 English‐speaking children followed longitudinally from Grade 3 to Grade 4 (i.e., from 8 to 9 years of age on average). At Grade 3, children read stories containing new words and answered questions about the spelling and meaning of these new words, evaluating orthographic and semantic learning, respectively. Children also completed outcome measures of orthographic representations (with a choice task targeting the spelling of existing words), word reading and reading comprehension (with standardised tasks) at Grades 3 and 4. We conducted regression analyses controlling for age, nonverbal reasoning, working memory, vocabulary and phonological awareness. Results We found that each of orthographic and semantic learning predicted gains in orthographic representations from Grade 3 to Grade 4. Furthermore, orthographic learning at Grade 3 predicted word reading at Grade 4, while semantic learning at Grade 3 predicted reading comprehension at Grade 4. Conclusions These longitudinal associations between orthographic and semantic learning and core aspects of reading strengthen the evidence in support of the hypothesis that children's word learning capacity plays a key role in reading development.

Beta vulgaris subsp. maritima (L.) Arcang. and Beta macrocarpa Guss. are crop wild relative taxa belonging to the primary gene pool. They constitute a crucial gene reserve for enhancing cultivated Beta species (B. vulgaris subsp. vulgaris L.). Climate change poses a significant threat to genetic reservoir in Tunisia. We evaluated the morphological diversity of ten populations of B. vulgaris subsp. maritima and five populations of B. macrocarpa growing in different Tunisian bioclimatic and ecological areas using a set of 9 quantitative and 14 qualitative traits to promote the preservation and exploration of this germplasm. Variance component analysis of the quantitative data showed an important spectrum of variability, both within and between populations. The principal component analysis (PCA) allocated this wild Beta collection into three groups. G1 included the populations of B. macrocarpa that were characterized by the largest glomerules and heaviest seeds, while G2 included all B. vulgaris subsp. maritima populations except one, i.e., N1015 that clustered into G3, which was characterized by the highest values of leaf characters. Similarly, qualitative traits exhibited a high diversity level (H'index ≥0.6) for almost all characters. The PCA divided these 15 populations into three groups as well: G′1 concerned the island B. vulgaris subsp. maritima populations, characterized by prostrate growth habit and red inflorescences; G′2 included all B. macrocarpa populations characterized by erect-procumbent growth habit and very synchronous flowering pattern; and G′3 was formed by the mainland B. vulgaris subsp. maritima populations, characterized by erect growth habit and hairy, curly leaves. The observed eco-geographic distribution patterns suggest that these wild relatives are highly adaptable to diverse and even extreme conditions (salinity, heat, and drought), highlighting their potential as resilient gene sources for beet breeding under the challenges of accelerating climate change.

Emotions are caused by a human brain reaction to objective events. The purpose of this study is to investigate emotion identification by machine learning using electroencephalography (EEG) data. Current research in EEG-based emotion recognition faces significant challenges due to the high-dimensionality and variability of EEG signals, which complicate accurate classification. Traditional methods often struggle to extract relevant features from noisy and high-dimensional data, and they typically fail to capture the complex temporal dependencies within EEG signals. Recent progress in machine learning by deep neural networks has opened up opportunities to develop methods highly efficient and practicable as to serve useful real-world applications. The purpose of this study is to investigate a novel end-to-end deep learning method of emotion recognition using EEG data, which prefaces a combination of two-dimensional (2D) convolutional network (CNN) and Long short-term memory network (LSTM) by an autoencoder. The autoencoder layers seek a lower dimensionality encoding for optimal input signal reconstruction, and the 2D CNN/LSTM combination layers capture both spatial and temporal features that best describe the emotion classes present in the data. Experiments in four-category classification of emotions, using the public and freely available DEAP dataset, revealed that the method reached superior performance: 90.04% for the "arousal" category, 89.97% for "valence", 87.73% for "dominance," and 90.84% for liking", as measured by the accuracy metric.

Federated Learning (FL) has gained attention across various industries for its capability to train machine learning models without centralizing sensitive data. While this approach offers significant benefits such as privacy preservation and decreased communication overhead, it presents several challenges, including deployment complexity and interoperability issues, particularly in heterogeneous scenarios or resource-constrained environments. Over-the-air (OTA) FL was introduced to tackle these challenges by disseminating model updates without necessitating direct device-to-device connections or centralized servers. However, OTA-FL brought forth limitations associated with heightened energy consumption and network latency. In this paper, we propose a multi-attribute client selection framework employing the grey wolf optimizer (GWO) to strategically control the number of participants in each round and optimize the OTA-FL process while considering accuracy, energy, delay, reliability, and fairness constraints of participating devices. We evaluate the performance of our multi-attribute client selection approach in terms of model loss minimization, convergence time reduction, and energy efficiency. In our experimental evaluation, we assessed and compared the performance of our approach against the existing state-of-the-art methods. Our results demonstrate that the proposed GWO-based client selection outperforms these baselines across various metrics. Specifically, our approach achieves a notable reduction in model loss, accelerates convergence time, and enhances energy efficiency while maintaining high fairness and reliability indicators.

The adoption of the Industrial Internet of Things (IIoT) in industries necessitates advancements in energy efficiency and latency reduction, especially for resource-constrained devices. Services require specific Quality of Service (QoS) levels to function properly, and meeting a threshold QoS can be sufficient for smooth connectivity, reducing the need to maximize perceived QoS due to energy concerns. This is modeled as a satisfactory game, aiming to find minimal power allocation to meet target demands. Due to environmental uncertainties, achieving a Robust Satisfactory Equilibrium (RSE) can be challenging, leading to less satisfaction. We propose a fully distributed, environment-aware power control scheme to enhance satisfaction in dynamic environments. The proposed Robust Banach-Picard (RBP) learning scheme combines deep learning and federated learning to overcome channel and interference impacts and accelerate convergence. Extensive simulations evaluate the scheme under varying channel states and QoS demands, with discussions on convergence speed, energy efficiency, scalability, complexity, and violation rate.

Several methods combining biomedical and computer-based approaches have been used to address the risk of falls among the elderly using instrumented insoles. Machine-learning techniques in gait analysis has proven to be a promising solution when using instrumented insoles. However, no study has investigated the risk of falls associated with a sequence of activities. Indeed, it can be observed that an important amount of energy is required by individuals preparing to get out of bed or toilet. The main goal of this work is to detect and associate different risk levels by analyzing the sit-to-start-of-walk (STSOW) sequence. Data were acquired during a Timed Up and Go test using an instrumented insole. The proposed approach compares six types of classifiers to the STSOW sequence signals. Then, a recursive clustering approach based on statistical features and the Kruskal Wallis test was implemented to define different levels of risk. The results show the capacity of the proposed approach to associate different risk levels of falls to an STSOW sequence. The accuracies of the classifiers ranged from 69% to 95.2%, and the best accuracy was achieved using both decision tree and ensemble classifiers. For the sit-to-start of the walk sequence identification phase, the best accuracy was achieved using the support vector machine model.

This study explored the reasons underlying pregnant women’s reluctance to undergo cesarean sections in Togo, despite its importance in reducing maternal and neonatal mortality. A total of 397 pregnant women who expressed hesitancy toward cesarean sections were enrolled during routine prenatal care visits at the country’s largest hospital. They completed a questionnaire comprising 72 statements addressing potential reasons for hesitancy toward cesarean sections. Their responses were analyzed using factor analysis, and the effects of participants’ demographic characteristics on scores for each factor were assessed using ANOVA. A seven-factor structure of motives was found: Fear of Death (endorsed by 92% of the sample); Regaining Autonomy Quickly (87%); Financial Concerns (74%); Fear of Stigmatization (73%); Fear of Unsupportive Reactions from Spouses and Relatives (72%); Prevention Through Spiritual Interventions (70%); and Perceived Health Risks for the Mother and Baby (40%). Scores on these factors were related to participants’ sociodemographic characteristics. Effectively addressing the low uptake of cesarean section requires a multifaceted approach rather than one focused on a single barrier. Our findings suggest critical points that could help develop tailored interventions to address the various obstacles to this life-saving care.

Background Imbalanced datasets pose challenges for developing accurate seizure detection systems based on electroencephalogram (EEG) data. Generative AI techniques may help augment minority class data to facilitate automatic epileptic seizure detection. New method This study investigates the impact of various data augmentation (DA) approaches, including Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP), Vanilla GAN, Conditional GAN (CGAN), and Cramer GAN, on classification performance with Random Forest models. The best-performing GAN variant, WGAN-GP, was then integrated with a bidirectional Long Short-Term Memory (LSTM) architecture and compared against traditional and synthetic oversampling methods. Results The evaluation of different GAN variants for data augmentation with Random Forest classifiers identified WGAN-GP as the most effective approach. The integration of WGAN-GP with bidirectional LSTM yielded substantial performance improvements, outperforming traditional oversampling methods and achieving an accuracy of 91.73% on the augmented data, compared to 86% accuracy on real data without augmentation. Comparison with existing methods The proposed generative AI approach combining WGAN-GP and recurrent neural network models outperforms comparative synthetic oversampling methods on metrics relevant for reliable seizure detection from imbalanced EEG datasets. Conclusions Incorporating the WGAN-GP generative AI technique for data augmentation and integrating it with bidirectional LSTM elevates seizure detection accuracy for imbalanced EEG datasets, surpassing the performance of traditional oversampling and class weight adjustment methods. This approach shows promise for improving epilepsy monitoring and management through enhanced automated detection system effectiveness.

Modern processors have instructions to process 16 bytes or more at once. These instructions are called SIMD, for single instruction, multiple data. Recent advances have leveraged SIMD instructions to accelerate parsing of common Internet formats such as JSON and base64. During HTML parsing, they quickly identify specific characters with a strategy called vectorized classification. We review their techniques and compare them with a faster alternative. We measure a 20-fold performance improvement in HTML scanning compared to traditional methods on recent ARM processors. Our findings highlight the potential of SIMD-based algorithms for optimizing Web browser performance.

This study explored the relationships between sexual health indicators (i.e., sexual satisfaction, distress, and function) and the DSM-5 Alternative Model for Personality Disorders, a promising dimensional framework for assessing personality pathology. A sample of 489 participants seeking help in private practice clinics completed self-report measures of sexual satisfaction, distress, and function, as well as dyadic adjustment, psychological distress, romantic attachment, personality impairment, and pathological personality facets. Results first showed that participants reaching the cutoffs for a personality disorder had significantly higher sexual distress and lower sexual function compared to participants without a personality disorder. Second, path analyses controlling for psychological distress, dyadic adjustment, and romantic attachment revealed that, for women, the Criterion B Intimacy Avoidance facet was consistently linked with higher sexual distress and lower sexual satisfaction and function. For men, result patterns were more complex, linking Criterion B Separation Insecurity with high sexual distress, and Separation Insecurity and Intimacy Avoidance facets with low sexual function. In addition, Criterion B Irresponsibility, Rigid Perfectionism, as well as Criterion A Intimacy impairment were linked with higher sexual satisfaction. These findings improve our understanding of the links between personality and sexual health and provide support for considering personality difficulties in sexual health interventions.

Background Characterizing the condition of patients suffering from knee osteoarthritis is complex due to multiple associations between clinical, functional, and structural parameters. While significant variability exists within this population, especially in candidates for total knee arthroplasty, there is increasing interest in knee kinematics among orthopedic surgeons aiming for more personalized approaches to achieve better outcomes and satisfaction. The primary objective of this study was to identify distinct kinematic phenotypes in total knee arthroplasty candidates and to compare different methods for the identification of these phenotypes. Methods Three-dimensional kinematic data obtained from a Knee Kinesiography exam during treadmill walking in the clinic were used. Various aspects of the clustering process were evaluated and compared to achieve optimal clustering, including data preparation, transformation, and representation methods. Results A K-Means clustering algorithm, performed using Euclidean distance, combined with principal component analysis applied on data transformed by standardization, was the optimal approach. Two unique kinematic phenotypes were identified among 80 total knee arthroplasty candidates. The two distinct phenotypes divided patients who significantly differed both in terms of knee kinematic representation and clinical outcomes, including a notable variation in 63.3% of frontal plane features and 81.8% of transverse plane features across 77.33% of the gait cycle, as well as differences in the Pain Catastrophizing Scale, highlighting the impact of these kinematic variations on patient pain and function. Conclusion Results from this study provide valuable insights for clinicians to develop personalized treatment approaches based on patients’ phenotype affiliation, ultimately helping to improve total knee arthroplasty outcomes.

Pseudorandom values are often generated as 64-bit binary words. These random words need to be converted into ranged values without statistical bias. We present an efficient algorithm to generate multiple independent uniformly-random bounded integers from a single uniformly-random binary word, without any bias. In the common case, our method uses one multiplication and no division operations per value produced. In practice, our algorithm can more than double the speed of unbiased random shuffling for small to moderately large arrays.

How does agency emerge eventfully in processes of organizational becoming? This article aims to address this question by developing a process theory of agency based on Gilbert Simondon's philosophical writings on individuation as a communicative phenomenon and Brian Massumi's writings on affect. This theory views agency as an affective force, expressed as a communicative event, that governs the transition from one process of individuation to another, producing an enhanced ability to act and potentially leading to a collective process of transindividuation that is essential to organizational becoming. In turn, this article not only offers novel theoretical as well as methodological insights for organizational research, but also highlights researchers' ethical responsibilities toward those whose individuation is precarious and who may not be able to partake in certain processes of organizational becoming.